atlas

Research map / Video / Video segmentation and tracking

Real-Time Neural Style Transfer for Videos

CVPR 2017 · Video › Video segmentation and tracking · approach: Self-supervised video representation learning · 236 citations (OpenAlex)

Check your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Train the stylizer with a temporal consistency loss rather than applying image style transfer independently to framesFeed-forward video style transfer · Sintel videostemporal error: 0.0770 -> 0.0439; 0.0926 -> 0.0675; 0.0517 -> 0.0304; 0.0789 -> 0.0553; 0.0872 -> 0.0513Helped
Use feed-forward stylization instead of per-frame optimizationVideo style transfer · Video framestemporal error: 0.0439 vs 0.0252; 0.0675 vs 0.0512; 0.0304 vs 0.0195; 0.0553 vs 0.0407; 0.0513 vs 0.0361Hurt
Add non-consecutive frame pairs during training (ablation)Ours · Sintel videostemporal error: 0.0250 -> 0.0241; 0.0406 -> 0.0394; 0.0182 -> 0.0180; 0.0330 -> 0.0311; 0.0297 -> 0.0283Helped
Increase residual-block feature maps from 48 to 128 (ablation)Ours · Sintel videostemporal error: 0.0244 -> 0.0243; 0.0425 -> 0.0408; 0.0195 -> 0.0193; 0.0334 -> 0.0330; 0.0302 -> 0.0296Mixed
Reduce residual-block feature maps from 128 to 48 (ablation)Ours · 1024 × 436 frame inferenceinference time: 0.098 seconds -> 0.041 secondsHelped

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy